Re: [microblaze-uclinux] [PATCH] return from interrupt.

[John Williams <jwilliams@xxxxxxxxxxxxxx>]

Hi Kenneth,


Kenneth Johansson wrote:
> It was possible to get a process stuck in an infinite loop on the rfid
> instruction. 
>
> The problem was that it was possible to get an interrupt during context
> restore and when that happened after R14 was restored it resulted in an
> infinite loop that gradually shrink down to include just the rfid
> instruction. 

That's funny timing, I made exactly the same patch to my tree yesterday 
while I was chasing a related bug!  It will be in the next release.

Thanks,

John

> Index: linux-2.6.x-petalogix/arch/microblaze/kernel/entry.S
> ===================================================================
> --- linux-2.6.x-petalogix.orig/arch/microblaze/kernel/entry.S   2007-11-12 11:42:12.000000000 +0100
> +++ linux-2.6.x-petalogix/arch/microblaze/kernel/entry.S        2007-11-12 11:43:44.000000000 +0100
> @@ -143,7 +143,11 @@
>  no_intr_reshed:
>         /* save mode indicator */
>         lwi     r11, r1, PT_MODE
> -3:     
> +3:
> +       /* Must protect the restore against interrupts              */
> +       /* schedule uses spin_lock_irq() and spin_unlock_irq()      */
> +       /* and even if schedule is not visited interrupts can be on */
> +       disable_irq     
>         swi     r11, r0, PER_CPU(KM)
>  
>         /* save r31 */
>
>
> ___________________________
> microblaze-uclinux mailing list
> microblaze-uclinux@xxxxxxxxxxxxxx
> Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
> Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/

___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/